Enhancement of guinea pig cytomegalovirus infection by two endogenously expressed components of the pentameric glycoprotein complex in epithelial cells

A better understanding of the mechanisms underlying cell tropisms and the efficiency of viral infection is critical for the development of vaccines and antiviral drugs for viral diseases. In this study, we worked on the entry mechanisms of guinea pig cytomegalovirus and found that endogenous expression of a combination of two components (GP131 and GP133) of the pentameric glycoprotein complex, which is required for non-fibroblast cell tropisms, enhanced viral infection more than 10-fold. In addition, D138A alteration in GP131 increased this enhancement by an additional 10-fold. Although differences in the efficiency of viral infection among various cell types are usually explained by differences in viral entry or traffic processes, our experimental evidences dismissed such possibilities. Instead, our findings that i) endogenous expression of GP131 and GP133 after nuclear delivery of viral DNA still enhanced infection and ii) an HDAC inhibitor overcame the need of the endogenous expression led us to hypothesize a novel mechanism that controls the efficiency of viral infection through the activation of gene expression from viral DNA delivered to the nuclei. Further studies of this unexpected phenomena warrant to understand novel but also general mechanisms for cell tropisms of viral infection and determinants that control infection efficiency.

the enhancement of infection is not due to increased endocytosis. To examine whether the co-expression of GP131 and GP133 increases endocytotic pathway activity in the absence of infection, FITC-conjugated Cholera toxin B subunit (CTB) or 70-kDa dextran was applied to the GP131/GP133-expressing cells or to the LacZ-expressing cells, and the cellular traffic of the FITC-conjugated substances from the plasma membrane to near the nuclei in the cells was monitored. Picture images show examples of the traffic of these FITC-conjugated substances. There were no significant differences in the traffic speed of FITC-CTB (Fig. 2a) or FITC-dextran (Fig. 2b) between the two conditions; i.e., in rAd-LacZ-and rAd-131/rAd-133-transduced cells, suggesting that GP131/GP133 co-expression has no effect on caveolae-dependent endocytosis or macropinocytosis. centrifugation immediately after infection additively improves infectivity. It is well known that slow centrifugation immediately after infection enhances CMV infectivity, although the mechanism for this remains unclear 22 . As shown in Fig. 3a, a 3-to 5-fold increase in infectivity was observed by 30 min-centrifugation immediately after infection in the cells transduced with rAd-LacZ, rAd-131, or a combination of rAd-131 and rAd-133, suggesting that the mechanism of the GP131/GP133-mediated enhancement differs from that of the centrifugation-mediated enhancement.

The enhancement depends on infection with virions but not with viral DNA. To investigate
whether the GP131/GP133-mediated enhancement of infection can also be initiated by the delivery of viral DNA instead of virions, BAC DNA containing the GPCMV WT genome (WT) was transfected into the cells transduced with rAd-LacZ, rAd-131, or a combination of rAd-131 and rAd-133. However, no significant differences in the numbers of the GFP-positive cells were observed (Fig. 3b). In addition, transfection of pBAC-GPCMV lacking GP129 expression (Δ129) resulted in no significant differences in the numbers of the GFP-positive cells between the cells transduced with rAd-131 and those with a combination of rAd-131 and rAd-133 (Fig. 3b). Combined with the fact that the GP131 gene product is incorporated in the virions 19 and the requirement for the GP131 gene in the GPCMV genome (Fig. 1c), these results led us to speculate that the engagement of cellular receptors with virions is required for the GP131/GP133-mediated enhancement of infection.
www.nature.com/scientificreports www.nature.com/scientificreports/ the enhancement is not due to an increase in viral attachment. The efficiencies of viral attachment to the GPE-7 cells expressing various combinations of the Pentamer components were compared. For this purpose, bromodeoxyuridine (BrdU)-labeled GPCMV WT stocks were prepared. Chondroitin sulfate A (CsA), an attachment inhibitor, was added to demonstrate the assay specificity. Although the average numbers of BrdU-labeled GPCMV particles attached to the cells transduced with rAd-LacZ, rAd-131, or rAd-133 were similar, those attached to the cells transduced with the combinations containing both rAd-131 and rAd-133 were reduced rather than increased (Fig. 4). Although the reduction was statistically significant (one way ANOVA, p < 0.001), we did not pursue the mechanism behind this observation. the enhancement remains dependent on endocytosis. As no effects were observed on the viral attachment processes for the co-expression of GP131 and GP133, next we analyzed the effects of their co-expression on the viral internalization process; i.e., endocytosis, during infection. Intracellular traffic of DiO-labeled GPCMV in the LacZ-and GP131/GP133-expressing cells was monitored. As DiO is a fluorescent probe for membrane, its signal with a certain mass can be detectable until its diffusion, i.e. at the point of fusion (a-c) GPE-7 cells were plated at 2.0 × 10 4 cells/well in 96-well plates. Next day, the cells were transduced with rAd-LacZ or with the indicated combinations of rAds encoding a Pentamer component (gH, gL, GP129, GP131, and GP133) at a total MOI of 100. Twenty-four hrs after rAd transduction, the cells were infected with GFP-expressing GPCMV WT (a,b) or Δ131 (c) at an MOI of 3 (based on titers determined in GPL cells). Two hrs later, the inoculum was replaced with fresh media, and the cells were incubated for additional 48 hrs. The numbers of GFP-positive cells were counted by taking digital images (a,c). The cells were also immunostained with monoclonal antibody g-1, which reacts with the GPCMV early antigen, and numbers of g-1-positive cells were counted (b). Means and standard deviations (SDs) of numbers in triplicate wells for each condition are shown. (d,e) Dose response of GP131/GP133-mediated enhancement. GPE-7 cells were plated and transduced with combinations of rAds encoding LacZ, GP131, and GP133 at the indicated MOIs. Twenty-four hrs later, the cells were infected with GPCMV WT and incubated for 48 hrs. The numbers of GFP-positive cells were counted, as described for (a). Means and SDs of numbers of GFP-positive cells in triplicate wells are shown. (f) GPC-16 cells were transduced with rAd-131 or with a combination of rAd-131 and rAd-133, infected with GPCMV WT, and incubated as described for GPE-7 cells in Panel a. Means and SDs of numbers of GFP-positive cells in triplicate wells for each condition are shown. of viral envelope with cellular membranes, in a confocal microscopic analysis. In other words, we expected that decay of DiO signals represented the traffic speed of virions from entry to membrane fusion. To make sure the specificity of the assay, first, cells were reacted with a free form of DiO of the amount equal to those in the DiO-labeled GPCMV stock used for infection. Any DiO signals were undetectable as dots in confocal microscopic analyses (Fig. 5a). In addition, CsA treatment of the cells significantly reduced the number of DiO signals in the DiO-GPCMV-infected cells to 15% of those without the treatment (Fig. 5a, time 0 hr). Both observations suggest that DiO signals observed are GPCMV-specific. At any time points during the first 2 hrs after shift-up to 37 o C, there were no significant differences in the traffic speed of GPCMV virions in the cells irrespective of GP131/GP133 expression (Fig. 5a). The effects of co-expression of GP131 and GP133 on endocytosis were examined by monitoring the presence of DiO-labeled GPCMV inside the cells after infection in the presence of genistein, dynasore, or latrunculin A (Fig. 5b). A broad-spectrum tyrosine kinase inhibitor genistein and a dynamin GTPase inhibitor dynasore interfere with caveolae-and clathrin-mediated endocytosis, respectively. In contrast, an actin polymerization inhibitor latrunculin A interferes with many other endocytotic mechanisms, including macropinocytosis 23 . Irrespective of the co-expression, latrunculin A decreased the DiO signals (Welch's t-test, p < 0.05), while genistein and dynasore had no significant effects. These observations suggest that GPCMV infection is dependent on a caveolae-and clathrin-independent endocytotic mechanism. However, there was no difference in the effect of latrunculin A between the cells transduced with rAd-LacZ and those with a combination of rAd-131 and rAd-133. These findings suggest that the GP131/GP133 co-expression did not change the nature of the endocytosis-dependent GPCMV infection.
Amino acid sequences of GP131 required for the enhancement of infection. rAds expressing GP131 with a charged amino acid-to-alanine alteration prepared previously 19 were used to identify the amino acid sequences required for the enhancement of infection. The rAd-mediated expression levels of all mutated GP131 proteins in human epithelial cell line HEK-293A were similar, while that of R152A was slightly less than www.nature.com/scientificreports www.nature.com/scientificreports/ those of the others and the mobilities of E47A and D51A in SDS-PAGE were slower than those of the wild-type (WT) and other mutated versions 19 . As shown in Fig. 6a, the E47A and D51A alterations abolished, and the K33A and K96A alterations reduced the enhancement of infection. Interestingly, the D138A alteration enhanced the GPCMV infection by an additional 10-fold. There were no apparent correlations between the enhancement and the requirements for infection that we reported previously 19 (Fig. 6b). As the amounts of GP131 WT and mutated versions were similar in HEK-293A cells transduced with their rAds 19 , we assume that the differences in the folds of enhancement were not due to different amounts of GP131 in cells. As the amounts of GP131 and other mutated versions in GPE-7 cells transduced with their rAds were marginal to be detected in immunoblotting and immunofluorescent assays, we could not compare them in GPE-7 cells.

Similar time course of appearance of Gfp and viral antigens irrespective of the enhancement.
As there were no differences in viral attachment or internalization to the cytoplasm, we next considered two possible scenarios. One potential scenario was that the nuclear delivery of viral DNA occurs more rapidly in the GP131/GP133-expressing cells. This scenario can expect time-dependent differences in the expression of GFP and viral antigens between the cells expressing LacZ and those expressing GP131/GP133. The other scenario  was that expression of the delivered viral DNA is more preferentially activated in the GP131/GP133-expressing cells than in the control cells. To distinguish between these two possibilities, the time course of the appearance of GFP-positive cells after GPCMV WT infection was compared among the cells transduced with rAd-LacZ, a combination of rAd-131 and rAd-133, and a combination of rAd expressing GP131 D138A (rAd-D138A) and rAd-133 ( Fig. 7). At 9~10 hrs after GPCMV WT infection, irrespective of GP131/GP133 expression, GFP-expression was almost undetectable under a microscope by the naked eye and only faintly detectable with the aid of a sensitive camera, but the number of GFP-positive cells was difficult to count. At 14 hrs post-infection, GFP-positive cells were detectable, but with some difficulty, while at 23 hrs post-infection, they were readily detectable. Thereafter, at 33 and 48 hrs post-infection, there was little increase in the number of GFP-positive cells. Although the number of GFP-positive cells differed among the cells transduced with the three combinations of rAds, the time course of the appearance and saturation of GFP-positive cells was similar for each, suggesting that the enhancement of infection is not due to time-dependent differences.
As CMV undergoes an ordered cascade of expressions of viral antigens, i.e. immediate early (IE), early (EA), and late (LA) gene antigens, upon infection of permissive cells, the appearance of cells expressing IE, EA or LA antigens was also compared among the cells transduced with rAd-LacZ, a combination of rAd-131 and rAd-133, and that of rAd-D138A and rAd-133 at indicated hrs after GPCMV WT infection. IE1/2 antigens were hardly detectable at 9 hrs, evident at 12 hrs, and clearly visible at 18 hrs after GPCMV infection among the cells transduced with the three combinations of rAds, confirming that the time course of GFP expression for each One day after transduction of GPE cells with rAds as described for (a), their culture media were replaced with pre-warmed medium containing the indicated inhibitors. The cells were incubated at 37 °C for 30 min and then placed on ice. Next, the cells were incubated with DiO-labeled GPCMV WT at an MOI of 22.5 on ice for 1 hr. After replacement of the inocula with pre-warmed medium containing the same inhibitor, the cells were incubated for 30 min or 1 hr, fixed, and treated as described for (a). Images were captured as described for (a), and means and SDs of relative DiO-positive signal numbers in three independent experiments are shown using the DiO-positive signal number in the cells that were transduced with rAd-LacZ and incubated without any inhibitor (none) as a 100% control. (2020) 10:8530 | https://doi.org/10.1038/s41598-020-65545-5 www.nature.com/scientificreports www.nature.com/scientificreports/  www.nature.com/scientificreports www.nature.com/scientificreports/ was similar to that of IE expression (Fig. 8). In addition, the EA antigen recognized by monoclonal antibody g-1 became faintly detectable at 12 hrs and clearly visible at 18 hrs post-infection. The early/late antigen gB 24 was weakly detectable at 48 hrs and strongly at 72 hrs post-infection. The number of cells expressing the EA or early/ late antigen was bigger in an order of the cells transduced with rAd-LacZ, a combination of rAd-131 and rAd-133, and that of rAd-D138A and rAd-133, indicating that all phases of viral infection were enhanced by the endogenous co-expression of GP131 and GP133.
To see if the differences in the number of GFP-or IE1/2-positive cells can be ascribed to the copy numbers of viral DNA in the nuclei, the average viral copy number in the nuclei per cell at 12 hrs post-infection was compared. There were no differences in viral loads (Fig. 9a), indicating that the endogenous co-expression of GP131 and GP133 has no effects on the viral nuclear delivery. To confirm whether viral genomes delivered to the nuclei have equal capabilities of viral gene expression irrespective of endogenous co-expression of GP131 and GP133, GPE-7 cells transduced with combinations of rAds were treated without or with trichostatin A (TSA), which selectively inhibits the class I and II mammalian histone deacetylase (HDAC) families of enzymes, for 12 hrs, as it is well known that HDAC is involved in repression of the IE1/2 gene promoter in non-permissive cells 25 . After the TSA treatment, the cells were infected with GPCMV WT at MOIs of 3 or 0.03, and the numbers of GFP-positive cells infected with GPCMV WT at an MOI of 3 were compared at 14 hrs post-infection (Fig. 9b). There were no apparent differences in the number of GFP-positive cells among the cells treated with TSA, and that number was significantly bigger than that of the TSA-untreated cells. In addition, the number of GPCMV IE1/2-positive cells was the same among the cells that were transduced with rAds, treated with TSA, and infected with GPCMV WT at an MOI of 0.03 (Fig. 9c), indicating that the viral DNA delivered to the nuclei has equal potential capabilities of viral gene expression.

Late expression of GP131/GP133 still enhanced infection. To see if co-expression of GP131 and
GP133 enhances viral gene expression from the viral DNA delivered to the nuclei, GPE-7 cells were transduced with combinations of rAds at 6 hrs and 12 hrs after GPCMV WT infection. As shown in Fig. 8, IE1/2 expression was already evident at 12 hrs after GPCMV infection. If co-expression of GP131 and GP133 increases viral DNA delivery to the nuclei physically, it is expected that their expression after viral DNA delivery may not increase GPCMV infection. However, co-expression of GP131 and GP133 enhanced GPCMV infection, the D138A mutation in GP131 showed additional enhancement, whereas the E47A and D51A mutations both abolished the enhancement (Fig. 10). Importantly, the time courses between the two time points of rAd transduction were similar. These results suggest that viral DNA delivered to the nuclei is more preferentially activated for expression in the GP131/GP133-and GP131 D138A/GP133-expressing cells.

Discussion
In this study, we found unexpectedly that expression of GP131 and GP133, two of the Pentamer components, enhanced the efficiency of GPCMV infection in epithelial cells. The enhancement was concluded not only based on the expression of GFP gene cloned into GPCMV WT, but also based on the expression of viral IE, EA, and early/late gene products. However, as shown in our previous study 19 , viral growth in GPE-7 cells was very slow, and it took at least 7 days to observe cell-to-cell spread, which makes hard to examine the effects on viral yields in a multiple step growth curve assay. We demonstrated that i) GP131/GP133 expression did not increase viral attachment or cellular virion traffic processes, and ii) endogenous co-expression of GP131 and GP133 did not compensate for the lack of GP131 in the GPCMV used for infection, which is different from trans-complementation of the defect in MCMV m131/m129 encoding Mck-2, a homolog of UL130 and GP131, to promote viral entry into macrophages as a gH/gL/MCK-2 complex 26 . We also demonstrated that the GP131/GP133-enhanced the infection www.nature.com/scientificreports www.nature.com/scientificreports/ of GPE-7 cells in an endocytosis-dependent manner as in normal infection. Previously, it was reported that bafilomycin treatment dramatically inhibited GPCMV infection of epithelial cells but did not greatly impact on fibroblast infection 18 . More vigorous studies are required to see whether the endocytotic pathway for the GPCMV infection of GPE-7 cells is similar to that for HCMV entry into some cell types [27][28][29] . Although it has been reported that HCMV is initially retained in early endosomes and then moves sequentially to the trans-Golgi network and  recycling endosomes before nuclear translocation in monocytes, and that the Src signaling pathway induced by viral attachment drives a unique nuclear translocation pattern 30,31 , this study was not intended to scrutinize such details. A comparison of the time course of GFP and viral antigen expression suggests that the enhancement of infection is not dependent on early increase of viral gene products expressed from the viral genome. Furthermore, enhancement of GPCMV infection, even by transduction with rAd-131 and rAd-133 at 12 hrs after GPCMV infection; i.e., after IE gene expression, supports the notion that the viral DNA delivered to the nuclei is more preferentially activated for viral IE-1/2 gene expression in the GP131/GP133-expressing cells. Also, an HDAC inhibitor overcame the need of the endogenous co-expression of GP131/GP133 for the increase of the numbers of the IE-1/2-expressing cells. Taking account of those findings, we hypothesize that expression of GP131/GP133 induces activation of viral gene expression from viral DNA delivered to the nuclei. However, we cannot exclude an alternative hypothesis that GP131/GP133 transduction influences release of viral genomes from capsids into the nucleus, as it was technically difficult to determine the copy numbers of nuclease-resistant viral genomic DNA. Although in the case of HCMV, it is reported that tumor necrosis factor alpha potentiates phosphorylation of KAP-1, a master co-repressor that silences IE gene expression by recruiting HP-1 and histone methyltransferase to the viral genome, for reactivation of virus from the latent myeloid cells 32,33 , our study found for the first time that endogenous expression of virion components activates GPCMV gene expression that is partly silenced in infected cells. It has been reported that induction of the HCMV signaling pathways for efficient infection depends on Pentamer components 30,34 . As GP131 and GP133 are gene products expressed at the late phase of infection, it is unlikely that their late expression induces such signaling pathways. However, as the precise roles of UL130 and UL131A, the HCMV homologs of GP131 and GP133, in the entry and signaling processes remain unclear, it is still possible to hypothesize that expression of GP131/GP133 induces the signal pathways for activation of viral gene expression that is achieved by the engagement of Pentamer with its receptors or by interactions with cellular molecules for signaling. Alternatively, endoplasmic reticulum stress induced by endogenous expression of GP131/ GP133 may activate viral gene expression through the unfolded protein response (UPR), as it has been documented that HCMV and MCMV infection, or their particular gene products, activates UPR and modulates viral gene expression 35,36 . However, such an interpretation could be inconsistent with our observation that the amounts of GP131 and its mutated versions with a single amino acid alteration were similar in the HEK-293A cells transduced with rAds encoding these proteins, while the degree of enhancement varied among the mutated versions.
To clarify the mechanism of the enhancement as well as the limitations of this study, it would be necessary to see (i) whether similar enhancement can be observed in HPV E6/E7-immortalized epithelial cells that show higher dependence on Pentamer for GPCMV entry 18 and in cell types other than epithelial cells, (ii) whether the enhancement is associated with immortalization of epithelial cells with SV40 T antigen that was reported to inhibit HCMV infection at multiple points in fibroblasts 37 and result in poor GPCMV infection 18 , (iii) the target molecules for GP131/GP133-mediated enhancement, and (iv) whether this enhancement reflects any process that occurs during the Pentamer-dependent signaling pathways. It is common in the field of virology that efficiency of infection in one cell type is different from that in another cell type. Although such differences are usually explained by differences in the viral entry or traffic process, this study suggests the presence of an additional mechanism that controls the efficiency of infection. Nevertheless, identification of the cellular targets remains critical for a better understanding this novel mechanism.
In conclusion, we identified an unexpected phenomenon in which endogenous co-expression of two components of Pentamer enhanced infection, probably through increase of IE-1/2 expression from the viral genome delivered to the nuclei. Although our findings represent an artificial enhancement of infection, we believe further studies to clarify the details may open a path to understanding the mechanisms of viral tropisms and determinants that control the efficiency of viral infection of cells.

Materials and Methods cells and viruses. GPCMVd9K (WT) expressing GFP and its mutants GP129Stop (Δ129) and GP131Stop
(Δ131), which contain an early stop-codon mutation in the GP129 and GP131 genes, respectively, were reported previously 17 . Production and titration of their viral stocks were performed in guinea pig fibroblast GPL cells (CCL158, ATCC). Although the infectivity depended on cell types, we used the GPL-based titers to calculate MOI. Guinea pig epithelial cell line GPE-7 was established previously by immortalization of primary renal cells with SV40 T antigen followed by cloning of cytokeratin-positive cells 19 . Infection of most GPE-7 cells with GPCMV required 100-fold amount of the viruses for infection of GPL cells. HEK-293A (Invitrogen) and the guinea pig epithelial cell line GPC-16 (CCL-242, ATCC) were purchased commercially.
The genes encoding the GPCMV Pentamer components, gH, gL, GP129, GP131, GP133, and GP131 with a charged amino acid-to-alanine alteration were cloned into pENTR3C (Invitrogen) respectively. rAds encoding each Pentamer component were constructed by the Gateway cloning from the pENTR3C constructs into pAd/ CMV/V5-DEST (Invitrogen) followed by transfection of PacI-digested plasmids into HEK-293A cells 19 . The rAd titers were measured in HEK-293A, and β-galactosidase staining of the cells transduced with rAd-LacZ indicated that infection at an MOI of 20 was required to transduce all GPE-7 cells.
immunological detection of GpcMV and its products. Polyclonal anti-GPCMV IE1/2 and anti-gB antibodies 13 and monoclonal antibody against a GPCMV early antigen g-1 21 were used for the immunological detection of GPCMV products. Immunostaining of infected cells in the wells of culture plates was performed as follows. Infected cells were fixed with 3.7% formalin, treated with 0.05% TritonX-100 in phosphate-buffered saline (PBS), rinsed with PBS, and incubated with a dilution of g-1, anti-GPCMV IE1/2 or anti-gB antibodies. After washing three times with PBS, the cells were incubated with horseradish peroxidase-conjugated anti-rabbit or mouse immunoglobulin G antibody (Histofine Simple Stain MAXPO, Nichirei, Japan) at 37 °C for 1 hr. Positive cells were visualized with 3,3′-diaminobenzidine (DAB substrate, Roche).
www.nature.com/scientificreports www.nature.com/scientificreports/ preparation and detection of BrdU-labeled GpcMV. One day after the infection of GPL cells with viral stocks, the infected cells were cultured in medium containing 10 µM BrdU (Sigma Aldrich). Cell-free BrdU-labeled virus stocks were prepared by the sucrose step gradient centrifugation of the culture supernatants of the infected cells. Infectious titers were measured in GPL cells. To detect BrdU-labeled GPCMV, the cells were grown and infected in 8-well LabTek chamber slides (glass) (ThermoFisher), and then the slides were treated with 4 M HCl for 10 min to denaturate DNA, rinsed with PBS, and reacted with anti-BrdU monoclonal antibody (3D4, Becton Dickinson) followed by FITC-conjugated rabbit anti-mouse IgG polyclonal antibody (DAKO). Evans blue was used for counter staining. After the reactions, BrdU signals on the plasma membrane were observed under a confocal microscope. Images of 20 fields (0.17 μm 2 /field) for each set of conditions were captured, and the virus signals were counted.
Preparation and detection of DiO-labeled GPCMV. DiO (Vybrant cell-labeling solution, Molecular Probe) was added to a GPCMV WT stock (4.1 × 10 7 PFUs in 180 μl of PBS) that was partially purified by sucrose gradient centrifugation 15 , at a final concentration 2 μM, and the mixture was incubated for 1 hr at room temperature. The mixture was used as the DiO-labeled GPCMV stock. GPE-7 cells in 8-well chamber slides (Nunc) were incubated with the DiO-labeled GPCMV at an MOI of 22.5 for 1 hr at 4 °C, and then the inoculum was replaced with warmed culture medium. The cells were incubated for the indicated duration at 37 °C, fixed with 3.7% formalin, stained with DAPI (NucBlue Fixed Cell Stain Ready Probes reagent, Life Technologies) for detection of nuclei and with 300-fold diluted acti-stain 555 fluorescent phalloidin (Cytoskeleton, Inc.) for detection of the polymerized form of F-actin, and observed under a confocal microscope. Images of 10 fields (0.17μm 2 /field) for each set of conditions were captured, and the virus signals were counted. fitc-labeled materials and inhibitors for endocytosis pathways. FITC-labeled-CTB, and -dextran (MW 70-kDa) were purchased from Sigma and used at final concentrations 10 μg/ml and 3 mg/ml, respectively. Inhibitors of endocytotic pathways, genistein (Wako, Japan), dynasore (Cayman Chemical, MI, US), and latrunculin (Wako), were used at final concentrations, 200, 50, and 2.5 μM, respectively.

Determination of viral loads. Preparation of the nuclei of the GPCMV-infected cells and purification of
DNA from the nuclei were performed as described previously for VZV-infected cells 38 . Quantitative PCR was performed as described previously 13 .

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.